Is there a part?

I didn't simulate it at all--there isn't any need. The requirements simply aren't that critical.

Gotcha! All mouth and no simulation (or breadboard) is the path to failure:

Simulated using an LED model based on actual LED V-I data taken by Paul Hovnanian...

com...

on the

to use the

=3D

in

hy

Perfectly fine -- 9 to 10.5mA over the battery voltage range, which is about what I estimated.

The user would never see it.

Compare, by contrast, to a simple resistor. That varies radically with everything, especially in the Vcc=3D3.3v case.

It looks like Paul's LED has a high Vf. Older GaN blue LEDs had Vf's around 3.3v, but now they're closer to 3v, likewise for greens.

(P.S. 10mA is much too much--it'll poke your eye out. 2mA is more like it these days. :-)

-- Cheers, James Arthur

m...

I calculated, in my head. It's not a complicated circuit. That's how I predicted the output variation.

I design first, then simulate only if needed.

-- Cheers, James Arthur

the

use the

The billboard drivers I'm designing go to 25mA... but they're truly flat over voltage, process and temperature (0.5% :-)

No one put numbers on anything, except for the 100-200mV thingy ;-)

Where did the thread go that wanted an indicator when battery voltage hit 4V? ...Jim Thompson

I design first, too, however I'm a perfectionist ;-) ...Jim Thompson

don't

I might, but that design is in the can. It works, though the LEDs do dim at the end of the battery's life.

Cost is an object. For less than a buck I could have thrown in a flyback LED driver.

=A0 =A0 =A0 =A0 =A0 =A0 =A0 ... Where I had headroom

covered at the

->

cost

change, or

added the

the

of just

was no going

case)

are

Nice. Hadn't seen (or thought up) that variation before. Funny luck of the draw. May even be able to use some very low threshold mosfet as = well.

at the

change, or

the

just

going

Actually, I did end up using a MOSFET (DGM1012T). ;-)

The neat thing is that when Vcc gets too low the MOSFET is turned on fully, so just adds Rds(on) to the ballast resistor. With a decent MOSFET, that's nothing, so it regulates the current when it can and doesn't hurt anything when it can't (down to where the MOSFET turns off).

the

For a solution, see...

Newsgroups: alt.binaries.schematics.electronic Subject: Re: Is there a part? - LED_Current_Source_SED.pdf Date: Sat, 04 Jun 2011 15:35:50 -0700 Message-ID:

Though, as I scratch my ass, I can see even a simpler mechanism. Can you ?:-) ...Jim Thompson

use the

Ass scratched version...

Newsgroups: alt.binaries.schematics.electronic Subject: Re: Is there a part? - LED_Current_Source_SED.pdf - LED_Current_Source_SED_2.pdf Date: Sat, 04 Jun 2011 16:56:44 -0700 Message-ID: ...Jim Thompson